Grantland: The Niners Are Running Wild

[Jikkle]

San Fran's balanced attack

I post it because it's a) a good read and b) from Bill Barnwell the same guy that posted we would decline based on some formula.

I'll give Barnwell credit in that he does pretty much own that his regression statement is looking to be very off. Guys like Pete Prisco over on CBS sports slowly and begrudgingly accept that they might have been off.

 

[Bemular]

San Fran's balanced attack

I post it because it's a) a good read and b) from Bill Barnwell the same guy that posted we would decline based on some formula.

I'll give Barnwell credit in that he does pretty much own that his regression statement is looking to be very off. Guys like Pete Prisco over on CBS sports slowly and begrudgingly accept that they might have been off.

Looks like he is using Football Outsiders stats/information - If I recall correctly FO had us winning the division with 8 or 9 wins and predicted our entire division would be horrible.

I think the NFC West is catching a lot of pundits with off guard.

 

[imac_21]

San Fran's balanced attack

I post it because it's a) a good read and b) from Bill Barnwell the same guy that posted we would decline based on some formula.

I'll give Barnwell credit in that he does pretty much own that his regression statement is looking to be very off. Guys like Pete Prisco over on CBS sports slowly and begrudgingly accept that they might have been off.

He didn't say we would be worse based on a formula, he said we would be worse based on history. So did Sando, Schatz, and many others.

If the over/under for Niner wins this year was set at 12.5 and you had to bet, would you take the over or the under?

 

[imac_21]

Looks like he is using Football Outsiders stats/information - If I recall correctly FO had us winning the division with 8 or 9 wins and predicted our entire division would be horrible.

I think the NFC West is catching a lot of pundits with off guard.

He wrote for FO before Grantland hired him. I believe he is credited with creating at least one of FO's "advanced stats."

 

[Bemular]

He wrote for FO before Grantland hired him. I believe he is credited with creating at least one of FO's "advanced stats."

Ahh, okay, very cool behind the scenes information - Thanks!

 

[supreme_clientele81]

San Fran's balanced attack

I'll give Barnwell credit in that he does pretty much own that his regression statement is looking to be very off.

I'll give him credit as well. Since the piece he wrote on the downfall of the 49ers he has pretty much willingly ate crow about so at least he is man enough to admit he was wrong.

And you can see from my signature what I thought about his "predictions"

 

[supreme_clientele81]

He didn't say we would be worse based on a formula, he said we would be worse based on history. So did Sando, Schatz, and many others.

If the over/under for Niner wins this year was set at 12.5 and you had to bet, would you take the over or the under?

But he took the under at 10 games. Most people would take the under at 12.5 (not me personally. I think we win 13 again) but under 10 was ridiculous IMO

 

[imac_21]

But he took the under at 10 games. Most people would take the under at 12.5 (not me personally. I think we win 13 again) but under 10 was ridiculous IMO

People taking the under at 12.5 are expecting regression.

As I said, he wasn't the only one that predicted. He was just the one with the biggest stage.

 

[mem49er]

People taking the under at 12.5 are expecting regression.

As I said, he wasn't the only one that predicted. He was just the one with the biggest stage.

But there is a significant difference between 12 wins and 9 wins, no?

 

[Flyingiguana]

these preseason predictions of a drop off are simply dumb writers going with the high probability.

 

[deep9er]

these preseason predictions of a drop off are simply dumb writers going with the high probability.

depends what you mean by "drop off"?

for me, still think a drop off in record will happen. teams are just tough all around, just have to look at West teams.

 

[Bemular]

depends what you mean by "drop off"?

for me, still think a drop off in record will happen. teams are just tough all around, just have to look at West teams.

To say a our record will drop off from 13 wins is nearly equal to saying the sun will rise tomorrow.

Over the past 10 seasons only 2.5 teams/Yr, on average, achieved 13+ wins. And, back-to-back 13+ win seasons has happened only twice in the past 10 seasons - both by NE. (2002-2003 & 2010-2011).

What won't drop off, however, is the level of our play - as we have already seen.

 

[spacedoodoopistol]

The problem with Barnwell's pre-season column is that it was a great example of how statheads put too much faith in stats as opposed to tangible factors. Stats are best as an "explanatory" device (what happened) rather than a "predictive" device (what will happen), but statheads get way overconfident about their predictive abilities.

He barely made any reference to why the Niners had made the improvement in the first place, or the actual reasons the team would fall short this year. It was just stuff about "well the 87 Bears and 92 Bills did so and so", which seemed entirely irrelevant to the question. It was just a stupid column.

 

[deep9er]

To say a our record will drop off from 13 wins is nearly equal to saying the sun will rise tomorrow.

Over the past 10 seasons only 2.5 teams/Yr, on average, achieved 13+ wins. And, back-to-back 13+ win seasons has happened only twice in the past 10 seasons - both by NE. (2002-2003 & 2010-2011).

What won't drop off, however, is the level of our play - as we have already seen.

yeah, the level of play will generally be what we've seen to now. but as in most sports, there are good days and bad days. we like to think Minnesota was the 'bad day', so we'll see?

as is always the case every year, the West games are key.

 

[Bemular]

The problem with Barnwell's pre-season column is that it was a great example of how statheads put too much faith in stats as opposed to tangible factors. Stats are best as an "explanatory" device (what happened) rather than a "predictive" device (what will happen), but statheads get way overconfident about their predictive abilities.

He barely made any reference to why the Niners had made the improvement in the first place, or the actual reasons the team would fall short this year. It was just stuff about "well the 87 Bears and 92 Bills did so and so", which seemed entirely irrelevant to the question. It was just a stupid column.

I recall my comment on the article suggested they (FO) failed to take into account that our "drop-off" year was 2010 from 2009. So, I agree with your sentiment on the Barnwell article.

But I have to disagree on your characterization of stats. Statistics are not at all an explanatory device - they explain nothing to be truthful. Statistics simply measure what happened. The explanation comes from understanding what happened.

I have seen all too often those who try to do this in reverse where they don't watch a game but they look at a box score and think they know just what happened in a game.

 

[Bemular]

yeah, the level of play will generally be what we've seen to now. but as in most sports, there are good days and bad days. we like to think Minnesota was the 'bad day', so we'll see?

as is always the case every year, the West games are key.

Keep in mind I am comparing this years skill and execution vs. last years. The drop-off is year over year not game over game.

 

[Jikkle]

The problem with using stats and history in terms of predicting football is that there is so many unquantifiable variables that go into whether you win or lose a football game.

Coaching and the chess match aspect of the game is one huge area that stats and history aren't going to predict. Even more so when it comes to us because a lot of our success and the past season and even this season is Harbaugh and staff's masterful ability to game plan and exploit weakness. Out thinking another team is never going to show up on a stat sheet.

 

[spacedoodoopistol]

I have seen all too often those who try to do this in reverse where they don't watch a game but they look at a box score and think they know just what happened in a game.

Definitely, this is the main thing that has gotten me bent about the new popularity of stats. People think they know everything because they have a few more stats. The only way to really understand a team or player is to watch and understand what you're seeing.

I think above I mean more like they're *better* to explain than predict. Definitely not all you need to explain though.

 

[imac_21]

But there is a significant difference between 12 wins and 9 wins, no?

Yes, but regression is regression. The guys saying the Niners would regress this year were using historical patterns to predict it. To not regress, they would have to win 13 games again.